Upload
others
View
2
Download
0
Embed Size (px)
Citation preview
DOCUMENT RESUME
ED 049 825 PS 004 531
AUTHORTITLE
SPONS AGENCY
PUB DATENOTE
Korner, Anneliese F.State as Variable, as Cbstacle and as Mediator ofStimulation in Infant Research.Stanford Univ, Calif. School of Medicine.National Inst. of Child Health and Human Development(NIH) , Bethesda, Md.12 Feb 7129p.; Pacer presented at the Merrill-PalmerConference on Research and Teaching of InfantDevelopment, Detroit, Michigan, February 12, 1971
EDES PRICE EDRS Price MF-$0.65 HC-$3.29DESCRIPTORS Behavior, hunger, *Infant Behavior, *Mediation
Theory, *Research Problems, Response Mode, Sleep,*Stimulation
IEENTIFIERS Physical State
ABSTRACTThis paper is a discussion of the different contexts
in whicn the concept cf the infant's state is used in infantresearch. The infant states discussed are: regular sleep, irregularsleep, drowsiness, alert inactivity, waking activity, and crying.Also included are hunger periods and indeterminate states, thoseinstances in which an infant's state does not clearly meet thecriteria of any of the other states. A brief summary of neonatalstudies which explore both innate and experiential factors affectingbehavior and development cf newborns is given. Without controllingfor state, infant studies may yield bcth false positive and falsenegative results. State can be considered a variable, the primaryfocus of research rather than an intervening factor. Research in thearea of state Froposed included those dealing with distinctness,predictability, range, and flexibility of state. State may also be amediator of stimulation. In studies assessing the effects of earlystimulation,, it is important to determine which types of stimulationare most effective at various developmental stages and the optimalstate of the infant during which such stimulation can take effect.Bibliography and tables included. (Authcr/AJ)
I-
U. k. DEPARTMENT OF HEALTH, EDUCATION & WELFAREOFFICE OF EDUCATION
THIS DOFII,VE:','T WiS BEEN 'PODUCED EXACTLY AS RECEIVED FROM THEPERSON OR ORG.-.NTTION 0'1-7.!ATING IT. POINTS OF VIEW OR OPINIONSSPIED DO NOT NECESSARILY REPRESENT OFFICIAL OFFICE OF EDUCATIONPOSITION OR POLICY.
STATE AS VARIABLE, AS OBSTACLE AND AS MEDIATOR OF STIMULATION IN
INFANT RESEARCH1
Anneliese F. Korner
Stanford University School of Medicine
In recent years, one of the most frequently discussed topics in the field
of infant research is the issue of the infant's state as it affects his behavior
and his responses. Most often, the infant's state is viewed as an obstacle, an
interference or a downright nuisance which, unless controlled for, stands in the
way of making reliable observations and of understanding the infant's responses.
Less frequently, state is discussed as a variable in its own right or as a neces-
sary pre-condition for stimulation to be effective. In this paper, I will attempt
to discuss separately the different contexts in which the concept of state is used,
at the risk perhaps, of artificially separating issues which overlap to some ex-
tent. In discussing and illustrating the contextual differences in which state is
used, I shall draw, as I was asked to do, mostly on examples from my own research
with newborns. On occasion, I shall also provide illustrations from the research
of other investigators.
When in our work with newborns we speak of the infants' states, we refer to
InPeter Wolff's (1966) classification of states. We use his behavioral criteria for
4714 determining state with only very minor modifications. Wolff's state definitions
correspond very closely to those described by Prechtl (1964) and his group of de-
velopmental neurologists in Groningen. Prechtl, in his highly sophisticated work
r) with neonates, probably was the first to stress the overriding and critical impor-
tance of considering state in evaluating any response on the part of the newborn.
1
Other investigators have devised slightly different classifications of the in-
fants' states. Among these are Brazelton's (1970), Janet Brown's (1964), Weller
and Bell's (1965) and most recently, Boismier's (1970) and his collaborators'
classifications. When one compares these different schemes, it is striking that
the overlap of criteria far exceeds the differences. This suggests that, even
though we are dealing with only a concept called state as opposed to a concrete
entity, there is substantive agreement regarding the criteria of state which are
useful in understanding an infant's behavior at any given point in time.
Let me briefly outline what states of the infant we will be talking about.
I shall do this in bold strokes and describe only the principal criteria by which
we identify the infant's state. For a more complete and detailed description of
the criteria used, I refer you to Wolff's monograph (1966) on the "Causes, Con-
trols and Organization of Behavior in the Newborn". We classify an infant as being
in .regular sleep when his eyes are closed and he is completely still for the most
part. His respirations are slow and regular; his face is relaxed, he does not
grimace and his eyelids are still. The infant is considered to be in irregular
sleep when his eyes are closed and when he engages in variable gentle limb move-
ments, writhing and stirring. Grimaces and other facial expressions are frequent;
respirations are irregular and faster than in regular sleep. Within this state,
interspersed and recurrent rapid eye movements can be seen through the eyelids.
These are a part of what the sleep researchers call the phasic events of paradoxi-
cal sleep or otherwise known as REM's. The infant is classified as being in the
state of drowsiness, when he is relatively inactive, when his eyes open and close
intermittently and when his respirations are fairly regular, though faster than
in regular sleep. When his eyes are open, they have a dull or glazed appearance.
This is in contrast to the appearance of his eyes during alert inactivity. In that
state, his eyes are open and have a bright and shining appearance. The infant is
capable in this state to pursue moving objects and to make conjugate eye movements
2
2
both in the horizontal and the vertical plane. The infant is relatively inac-
tive in this state; his face is relaxed and he does not grimace. In waking
activity, the infant frequently engages in diffuse motor activity involving his
whole body. His eyes are open but not alert. Respirations are grossly irregu-
lar. The state of crying is characterized by crying vocalizations associated
with vigorous, diffuse motor activity.
These states have often erroneously been called states of arousal. Un-
doubtedly, this came to pass because superficially they appear to express the
range from the infant's lowest to the highest degree of activation. As Wolff
has frequently pointed out, these states do not represent a continuum of arousal
of the infant, but are qualitatively different conditions, with a distinct type
of internal organization. This is very clearly born out by EEG criteria of various
states which show that the brain is at least as activated or "aroused" during
irregular REM sleep as during waking.
You may wonder, why we don't use EEG criteria altogether to determine state.
In some types of infant research this is absolutely essential, particularly in
certain sleep studies and in investigations primarily concerned with brain matura-
tion. Not being a psychophysiologist, I personally don't "hook up" babies, as it
is called in the trade. It is reassuring and of interest that the behavioral state
criteria correlate quite well with the EEG criteria. Conceivably, the behavioral
criteria may even be more accurate than the EEG criteria in describing the state of
the infant, because the behavioral criteria were devised with the infant in mind,
whereas the EEG criteria had to be converted and adapted from adult standards.
In my work with newborns I have added a state to those of Wolff's, a state
which I have loosely called an indeterminate state. This classification is used
in those instances in which the infant's state does not clearly meet the criteria
of any of the other states. This category was added not only to avoid misclassi-
fications but also as a behavioral variable in its own right. As I will try to
3
3
show later, the fact that some infants convey and perhaps experience their in-
ternal states less clearly than others may be an important attribute in their
make-up which may become a source of confusion to their caretaker and may delay
the development of internal sets of expectations.
In my discussion of the impact of state on neonatal behavior, I shall also
refer to the influence of hunger. While hunger is an entirely different phenom-
enon from the states already described, one may nevertheless think of hunger as
a state in terms of its periodic, universal and biological nature. Since there
is no reliable index of the intensity with which a given infant experiences hun-
ger, hunger is of necessity an inferential state. We thus inferred hunger from
the length of time elapsed since the last feeding, assuming that hunger would in-
crease as a function of time.
Before I will proceed to illustrate from my neonatal studies in what way
state can be an obstacle, a variable, or a mediator of stimulation, it might help
to give you first a very general description of these studies.
Description of Neonatal Studies
Our neonatal studies are designed to explore both innate and experiential
factors as these affect the behavior and development of newborns. We have long
been involved in studies of individual differences among neonates and have tried
to assess these differences before at least extra-uterine experience could much
affect the infants in a differential way. Our interest in assessing individual
differences has been governed by three very general considerations: We feel that
individual differences assessed shortly after birth may contain rudiments of the
individual's later characteristics and may differentially affect the unfolding of
many of his later functions. We also strongly suspect that such differences may
influence the manner in which universal childhood events are experienced and per-
ceived by any given child. Furthermore, we feel that such variations should, in
the interest of mother-infant mutuality, evoke differences in mothering.
4
4
In our early experience studies we have sought to identify the maturationally
most adequate forms of stimulation which are relevant for the first weeks of life.
We have attempted to pinpoint these in the context of the types of stimulation or-
dinarily provided by maternal care. For both types of studies, we have included
in our samples only full-term, healthy newborns. this was particularly important
in the individual differences studies since otherwise we might have found differ-
ences among babies purely as a function of pre- and postnatal complications. Thus
we applied stringent selection criteria with respect to maternal labor, medication
and delivery, parental metabolic and neurological status, the infant's condition
at birth, his weight range and his physical status during the entire lying-in peri-
od. This was done to make certain that only the healthiest babies were included
for study.
We developed standard procedures of observation for each of our studies. In
this way we observedthe infants under identical external conditions which insured
that whatever individual differences we found were not a function of differing
conditions of observation. We always observe the infants at a comparable time of
day. Illumination and temperature are kept constant. Since the position of the
infant frequently affects his responses, we control for this factor in each obser-
vation. In our experimental studies, the sensory stimuli provided are standard
in duration and intensity.
In one of our studies, we used film to record the infant's behavior. In
/:); fact, we have taken 1000 feet of film on each of 32 babies and we have thus accumu-,69
lated a bank of behavioral data on the neonate which can be used by us and other
investigators for a host of investigative purposes. Sixteen time samples were
taken on each infant, interspersed among two prefeed half hours, one postfeed
40 half hour and during the half hour at midpoint i-i.ween feedings. A timer attached
3,47 to the camera automatically turned the camera on and off, thus taking behavior sam-
ples which were identical for each infant in the length and in the interval since
5
the last feeding.
With the help of a computer attachment built for our projector which allows
us to analyze film at a predetermined number of frame units at a time, we have
to date undertaken five separate film analyses. These have revolved largely
around individual differences in various spontaneous oral behaviors, the frequen-
cy and the types of the infants' motions, and the relation between the infants'
oral and motility characteristics (Korner, Chuck and Dontchos, 1968; Korner and
Kraemer, 1970). Most recently, we have concentrated on exploring the temporal
relation between two of the most highly organized neonatal behaviors, namely
hand to mouth coordination and visual behavior.
Observer reliabilities, incedentally have been high throughout, both in
the direct observations of the infants and in the film analyses. At the begin-
ning of each study and periodically throughout we check inter-observer agreements.
Agreements have been way above 90% in all but two behavior categories. In the
film analyses, in order to avoid subtle changes in scoring criteria over time,
we not only establish inter-observer agreement levels, but also score - restore
reliabilities for the same person throughout the period film is analysed.
Now that I have described in a general way what kinds of studies we have
undertaken, let me turn to the discussion of the relevancy of state in these
studies. I shall begin by illustrating in what way state can be an obstacle
'n infant research.
State as Obstacle in Infant Research
Escalona (1962), in a classical paper entitled "The study of individual
differences and the problem of state" called the state variable a source of
"bedevilment" to investigators who attempt to study the behavior of young in-
fants. This is particularly true in studies of individual differences, as I will
try to illustrate from our studies. It is, however, also true of studies which
attempt to investigate what infants have in common at any given developmental
6
stage, or how they respond to differences in experimental treatments, inter-
estingly enough, if one does not contr for state, one is apt to obtain both
false positive and false negative results. False positive results are most
likely to occur in individual difference studies. Let me illustrate from the
influence state exerts on the infant's responsiveness to sensory stimulation.
In one of our studies we used a buzzer emitting unpure tone of approximately
80 decibels for 2/5 of a seconc at a distance of about 2 feet from the infant's
exposed ear. In a preliminary unpublished study, I found, as other investigators
have (i.e. Wolff, 1959; Eisenberg, Griffin, Coursin and Hunter, 1964) that thresh-
olds to auditory stimulation vary with the state the infant happens to be in, and
that they tend to be higher in regular sleep and during crying, than during irreg-
ular sleep, drowsiness or alert inactivity. I also noted, as Wolff (1959) already
had, that not only did the frequency of response as gauged by indices of behavioral
change vary with the infant's state, but also the kind of response the infant would
give in response to auditory stimulation. Thus, if he did respond at all during
regular sleep, he was apt to react with a vigorous startle, while the same stimu-
lus would only evoke a palpepral blink or a minor body movement during irregular
sleep. From this, it was quite clear that if we wished to compare infants with
respect to their relative responsiveness to auditory stimulation, they all had to
be tested in the same state, and as I will show later, during a comparable time
interval since the last feeding. In the case of auditory stimulation, we chose
to test all infants while they were in irregular sleep and this choice was based
on the rationale that irregular sleep is very prevalent among newborns and that
auditory thresholds in this state are relatively low. With other types of sensory
stimulation experiments we, of necessity, also had to choose a pre-determined,
pre-stimulus state, for only in so doing, were we able to compare infants among
each other with respect to their relative sensory sensitivity. Clearly, unless one
controls for state, one is apt to find false individual differences, differences
7
which are purely a function of the infant's state rather than his individuality.
False negative results are also easy to come by if one does not control for
state. These plague more frequently the types of studies which investigate the
effect of an experimental treatment as well as studies which explore what certain
populations of infants have in common. If, for example, in a conditioning study
one does not strictly control for the infant's state at the time of testing, the
infant's response may be so variable that one may conclude wrongly that the con-
ditioning had little or no effect (See a discussion of this problem by Connolly
and Stratton, 1969). Or, if one wants to identify neurologically suspect newborns
as Lenard and other associates of Prechtl's (1968) have done with admirable prog-
nostication, it is absolutely imperative that one test each of the reflexes while
the infants are in a comparable state. Otherwise, one may judge as abnormal the
absence of reflex responses, a failure which may be purely a function of the in-
fant's momentary state.
State does not necessarily have to represent a source of error or confusion.
If routinely taken into account, it is an aid in systematic observations, a refer-
ence point regarding the infant's behavior at any given moment, and an organizer
of otherwise confusing data. Any naive observer can testify how utterly confusing
it 17 to watch a newborn and to bring any sense or order to his observations. TheLS
sense of bewilderment vanishes when the infant's behaviors are brought into the
context of his states. For example, infants startle, grimace and smile, they suck,
they tongue and engage in rhythmical mo'Jching. Boy babies have erections. Some
of these spontaneous behaviors are not at all random, but they occur in each baby
only in certain states. Also, their frequency varies with each state. For exam-
ple, reflex smiles in the newborn never occur in regular sleep. They are most
common in irregular sleep with and without REM's. They occasionally occur during
drowsiness. On the rarest occasions, and usually only in response to a sudden vis-
ual and auditory stimulus and right after a feeding, do smiles occur during alert
8
8
inactivity. Vigorous startles are by far the most frequent uring regular sleep;
they are less common and less vigorous during irregular slee and drowsiness and
are extremely rare in the waking states. Erections occur prinarily during irreg-
ular sleep, particularly duriny the rapid eye movement phase. To a lesser extent,
they also occur in regular sleep and during drowsiness. Rhythmical mouthing is
primarily associated with regular sleep.
These and other findings were derived from systematic ob ervations of these
phenomena, first by Wolff (1966) and later confirmed in one o
I
our studies (Korner,
1
1969i. Through lengthy monitoring of states, we quantified d e hourly rate of each
of these spontaneous behaviors in each state. Interesting se; and individual dif-
ferences emerged in our study from these observations. The r sults on sex differ-
ences reflected a highly consistent trend for males to startl: more in all states
and for females to engage more frequently in reflex smiles and bursts of rhythmical
mouthing. S nce the mean rate of the spontaneous behaviors 1A3s almost identical
for males and females when erections were excluded, it seem& that females made up
in smiles and reflex sucks what they lacked in startles. It also appeared from our
results that individual infants heavily relied on specific cischarge channels over
states. Thus, infants who startled a lot in regular sleep slso tended to be the
ones who startled frequently in irregular sleep. The rank correlation was .49,
significant at p '. .01. This relation was more marked in he case of erections.
The rank correlation between the rate of erections in regu ar sleep and irregular
sleep was .77, significant at p G .001.
Applying these findings to the issue of state being either a confusing or an
organizing factor, it is clear that one may draw the wrong conclusions about the
individual characteristics of a given baby either from the frequent occurrence or
the total absence of these spontaneous behaviors, if one does not consider the
state of the infant. It is equally clear, that if one is not interested in indi-
vidual differences, but wants to investigate in a phenomenological way, the nature
9
9
and the prevalence of these spontaneous behaviors among babies, consideration of
the state in which these behaviors occur, is a most powerful ally in organizing
otherwise highly confusing observations.
So far, I have talked only about the influence of the sleeping and waking
states on the infant's behavior. I want to illustrate next how hunger, as inferred
by the time elapsed since the last feeding, may influence an infant's behavioral
response. We systematically gathered much of our data in the half hours immediately
following and preceding a feeding and at midpoint between feeds. Analyses of var-
iance showed highly significant differences between periods in a great number of
behaviors. For example, the infants' response to auditory stimulation differed
significantly between periods, with the infants being most responsive right after
a feeding, and least responsive midway between feedings. The frequency of mouthing
and of the rooting response increased significantly as feeding time approached. To-
tal movement, and particularly diffuse motions increased significantly with hunger,
and single motions decreased accordingly. In the state dimensions, sleeping states,
particularly regular sleep significantly decreased with hunger, whereas crying fre-
quency and the number of shifts of states increased. None of these findings are
particularly surprising or contrzry to expectations. They do however, convey a
point that must be considered by anyone doing infant research and that is, that in-
fants behave differently depending on when during the hunger cycle they are observed.
Without controlling for the time elapsed since the last feeding, one may once more
obscure group effects which actually may exist, or find individual differences which
are an artifact.
State as a Variable
State, as I hope to show, can be considered as a variable in its own right,
as the main focus of research rather than as an intervening factor which either
clarifies or obscures the meaningfulness of observations of other types of behavior.
State as variable has been one of the most and one of the least researched areas of
investigation. Let me begin with some of the least explored areas which pose sub-
10
10
tle and difficult problems for investigation but which, in my opinion, also hold
great promise in shedding light on both normal and pathological development.
One variable in this realm is the distinctness of the infant's state. In-
fants vary a great deal in the clarity with which they convey their state. As
mentioned before, we attempted to assess this phenomenon by adding a category
called "indeterminate state" to Wolff's (1966) state categories, which we used
whenever the criteria for classifying an infant's state were fuzzy or blurred.
We did not get very far in pursuing this problem because it was a side issue a-
midst many others at the time, and therefore was not assessed as thoroughly as it
would have 3een had it been the main focus of investigation. Nevertheless, I feel
that it would be a very important area to explore, for indistinctness of the mani-
festation of the infant's states, including those of hunger, undoubtedly represents
an individual's first failure to communicate. Indistinctness of state, great varia-
bility and unpredictability of states must have a profound effect on the beginning
mother-infant relationship. When mothers talk about easy or difficult babies to
take care of, they undoubtedly refer in great part to the ease with which they can
"read" their infant's needs at any given moment. Equally important, particularly
for the infant's later development, may bt the distinctness with which he experiences
the internal stimuli generated by his states, especially those of hunger. Conceiv-
ably, such indistinctness may delay the development of internal sets of expectations
and of the anticipatory functions. Blurredness of the experience of internal states
may also delay the distinctions between internal and external reality and may predis-
pose to later regression in this distinction.
Of interest in this connection are reports from psychopathology. For exam-
ple, adult schizophrenics show on EEG an unusually large percentage of "intermediary"
sleep stages which are so clas :,ified on the basis that they simultaneously show
characteristics of several sleep stages (Koresko, Snyder and Feinberg, 1963;
Vincent, Favarel-Garrigues, Bourgeois and Dufet, 1968). Also of interest are
reports both on obese patients and on patients
11
11
with anorexia nervosa to the effect that these patients have not learned to know
when they are hungry (Griggs and Stunkard, 1964; Bruch, 1966). Particularly
interesting in this connection is Dowling's report (1970) on the development of
infants who were never allowed to experience hunger. He described three children
born with esophageal atresia, a correctable, congenital anomaly which prevents
passage of food from mouth to stomach. Surgical repair of these particular chil-
dren was not possible until they were two years old. Until then, they received
nourishment through a gastric fistula which was made surgically through an opening
leading to the stomach. Food was given two of these children without oral partici-
pation on a 3 hourly schedule, day and night, asleep or awake, without relation to
hunger. In each of the children, an opening was also made from the upper segment
of the esophagus to the side of the neck. In this way, anything the infant swal-
lowed passed through this opening to the outside. The two children who experienced
no oral feedings simultaneously with the gastric feedings, soon developed gross de-
velopmental deficits. When one of these children was first seen at 4 weeks of age,
both his rooting and sucking reflexes were very weak. Neither of these two children
used the mouth, as normally developing infants do, for purposes of exploring, feel-
ing or grasping. Most remarkable was the fact that both these children had a marked
lag in gross motor development during the first 2 years of life, even though, on
testing both had normal and above normal intelligence at the age of 5 and 6 years.
The development of the third infant who was allowed to experience hunger and, who
when fed gastrically, also was given solids, was radically different. She was very
much interested in food, cried hungrily when waiting for it, becoming active and
eager during its preparation. At 22 months of age this child was vigorous, explor-
atory, and active, and before all, she did not show any of the gross motor lag seen
in the two other infants. In discussing the difference in development between these
three childr..,n, Dowling (1970) invoked the organizing and activation force of hunger
as one important factor underlying the lack of gross motor deficit seen in the more
12
12
normally developing child.
One may profitably study not only the distinctness and predictability of
states, but also the range and flexibility of the states which an infant is capa-
ble of. Brazelton (1962) observed several children, who, during the first post-
natal weeks, showed a very narrow range of states. These children came under
intensive and unsuccessful psychiatric care at a very early age and were later
diagnosed as children with "atypical development". Brazelton gave a detailed de-
scription of one of these children who essentially was capable of only two states:
in the first state the infant appeared to be in deep sleep during which his muscle
tone was poor, he was difficult to rouse and impervious to any stimulation. In
the second state, he continuously screamed, was hyperactive and hypersensitive to
any stimulation. Nothing could calm him except restraint and swaddling which made
him revert immediately back to the first state of inaccessible, deep sleep. Repeat-
ed neurological examinations revealed no deficit. His development progressed very
unevenly, particularly in the motor sphere. The narrow range of this child's states
persisted at least until pre-school age when his coping mechanisms were limited to
either screaming or withdrawing into a state in which he seemed to neither hear nor
see. Aside from the internal difficulties which these state peculiarities engen-
dered, they also had a profound effect on the mother-infant relationship. The
mother of this infant was completely overwhelmed by her inability to comfort or
reach this child and by his unresponsiveness to her ministrations.
In addition to distinctness, consistency and range of states, alterability of
state through stimulation would be a very important variable to study systematically.
For example, in comparable states, some infants are extremely difficult to rouse,
others respond with minimal stimulation. Similarly, some crying, agitated infants
are easily calmed by certain interventions, whereas others react minimally or only
very slowly to the very same methods of soothing (Bridger and Birns, 1963; Korner
and Thoman, in preparation). The study of the alterability of state may have much
13
13
larger applicability than either distinctness or range of states in that it may
be an important dimension in normal development, and exert a significant effect
on later characterological traits. For example, infants who maintain their re-
luctance to being roused may become phlegmatic, slow to respond, placid individuals;
by contrast, infants who respond to any kind of handling or stimulation with rapid
changes in state may become individuals who are highly responsive to any kind of
stimulation, to the point, perhaps, of becoming easily overstimulated.
Probably the best researched area in which state is treated as a variable is
that of sleep. The last ten years have seen a veritable avalanche of sleep research,
and this field has become, without a doubt, one of the most exciting new frontiers
in psychological investigation. A number of investigators have systematically ex-
plored the sleep characteristics of newborn infants (i.e. Roffwarg, Muzio and Dement,
1966; Ellingson, 1967; Monod, Eliet-Flescher and Dreyfus Brisac, 1967; Parmelee,
Wenner, Akiyama, Schultz and Stern, 1967; Prechtl, Akiyama, Zinkin and Grant, 1968;
Emde and Metcalf, 1970). One of the recurrent questions posed by this research
revolves around the function which the prevalence of the REM state might have in4/0 .hem
the developing organism. It is an established fact thati the younger the organism,
animal or human, the more prevalent the REM state (Parmelee et al, 1967, Jouvet-
Mounier, 1968). Roffwarg et al (1966) advanced the highly plausable hypothesis
that the great prominence of the REM state during early life may serve to assist
the process of central nervous system maturation and differentiation through the
endogenous afferent stimulation which the REM state provides. Within the frame-
work of my study on individual differences, I examined the behavioral manifestations
of the REM state, comparing in developmental terms, the organization of the infant's
REM state with that of the adult (Korner, 1968). It was of interest that, with one
notable exception, the comparison between infant and adult REM associated behaviors
highlighted the well known developmental principle which characterizes development
as proceeding from the general, or undifferentiated to the increasingly specific
14
14
and differentiated (Hamburger, 1947; Werner, 1947). Thus for example, in the
normal adult, REM onset is always preceded by 50-90 minutes of nonREM sleep.
This contrasts with REM onset in neonates who, in an indiscriminate, undiffer-
entiated way enter REM sleep from almost any state, waking or sleeping. This
had already been shown by Roffwarg and his collaborators (1966). In our sample,
only 25% of the REM episodes were preceded by regular sleep. The remaining REM
periods were preceded by drowsiness, crying and waking activity, in that order
of frequency. Similarly, erections which in normal adults occur almost exclu-
sively and synchronously with REM periods (Fisher, Gross and Zuch, 1965), were
neither as REM-specific nor as synchronous in newborns. While erections in neo-
nates do primarily occur during REM, which is an interesting precursor to the later
obligatory association, they also occur with considerable frequency during regular
sleep and to a lesser extent while the infant is drowsy. In each comparison, the
REM characteristics of neonates were more generalized and less specific as to func-
tion and temporal relations than the adults'. The one exception was in the area
of responsiveness to auditory stimulation during REM's and during other states.
We found responsiveness to be significantly less frequent during REM's than during
any other state except during "deep" or regular sleep. This suggests that during
the REM phase, the infant is as impervious to external stimulation as he is during
regular sleep, a finding which is apparently also true for adults (Williams, Hammack,
Daly, Dement and Lubin, 1964). What is fascinating also is that when one examines
the literature on adult sleep pathology, one finds under the impact of extreme stress,
a reversion or regression to earlier levels of REM organization in some of the char-
acteristics described. However, this is another long and complicated story, too
far removed from our main concerns.
Returning to our topic of state being used as a variable, still other illustra-
tions can be found in investigations focusing on the infant's crying or his capacity
for alert inactivity. In certain respects, some of these studies could also be
15
15
classified under the heading of state being a mediator, as well as an evoker of
:-.:imulation. For example, in our individual differences sLudy (Korner and Grob-
stein, 1967), we monitored the frequency and duration crying during 2 hours
and 20 minutes on each of 32 subjects. We found significant differences among
babies in how much they cried, and when they cried most during the four hour
feeding schedule which is prevalent in most hospitals. There were some infants
who cried most three hours after the beginning of the last feeding and others who
cried most just before the next scheduled feeding. This clearly suggests that some
infants would do better on a three hour schedule, particularly since these infants
often exhaust themselves crying and fall asleep just before they are scheduled to
feed. The individual differences in the amount of crying between babies also imply
the need for different care. A very irritable infant not only must affect the moth-
er and her feelings as a caretaker, but will also evoke all sorts of efforts on her
part. Since the infant's crying usually elicits maternal ministrations, differences
in crying will affect the frequency with which the infant initiate$ interactions
with his mother. In that sense, the state of crying may be considered a mediator or
evoker of stimulation.
Crying has also been taken as a variable in diagnostic studies (Fisichelli,
Haber, Davis and Karelitz, 1966; Prechtl, Theorell, Gramsbergen and Lind, 1969).
There appear to be marked qualitative differences between the cry of the brain
injured and the normal infant. In a similar vein, Brazelton and his collaborators
(1966) have taken the infant's capacity for the state of alert inactivity and for
visual following as a favorable prognostic sign for the neurological intactness of
the infant.
The state of alert inactivity has long captured our interest, in that the
visual exploratory behavior which often attends this state may be one of the main
avenues at the neonate's disposal for getting acquainted with his environment and
for early learning. In this sense, the state of alert inactivity is both a varia-
16
16
ble and a mediator of stimulation. In our studies, we have been interested both
in the general make-up of those infants who are capable of a great deal of visual
alertness and in identifying the types of interventions within the context of or-
dinary maternal care, which will evoke visual alertness in newborns most effectively.
In an attempt to find out what the characteristics are of those babies who are capa-
ble of a great deal of visual alertnes9, we monitored the frequency and duration of
alert inactivity and correlated these with our other findings on these babies. Among
the correlates, one of the most interesting was the strong relation between the state
of alert inactivity and the capacity for visual pursuit. The correlations between
visual pursuit and duration and frequency of alert inactivity were .60 and .74 re-
spectively (p < .001, Korner, 1970). Thus, the babies who spontaneously spent a
great deal of time in alert inactivity, also tended to be the most capable of fix-
ating on specific visual stimuli. Both measures thus express a capacity which may
be vital for the earliest forms of learning.
State as Mediator of Stimulation
Having already given two examples in which state could be considered not only
as a variable, but also as a mediator of stimulation, I shall limit myself to one
illustration which, hopefullly, will further clarify in what way state can be a medi-
ator of stimulation. Dr. Evelyn Thoman and I recently completed a study in which
we attempted to pinpoint what types of stimulation mothers ordinarily provide in
the course of caring for their infants (Korner and Thoman, 1970). In particular,
we assessed the relative efficacy of body contact and of vestibular stimulation
(the stimulation involved in being moved), in evoking visual alertness in newborns.
To make my point as to state being a mediator of stimulation, I will need to sum-
marize this study briefly. The subjects were 40 crying and 24 sleeping 2-3 day old
newborns, equally divided as to males and females, breast-fed and bottle-fed infants.
Each subject was given, in random order, six interventions which entailed singly,
or in combination, contact and vestibular stimulation with or without the upright
17
17
position. The highest level of alerting achieved during 30 seconds pre- and
post intervention was judged by two observers on a 6 point scale. In order to
avoid state becoming an obstacle to reliable observations, we controlled for pre-
stimulus state. Crying infants were tested only after- they fussed for at least
one minute and cried vigorously for 10 seconds. Sleeping infants were tested only
during irregular sleep without REM's. To avoid experiment-induced heightened
arousal, at least two minutes had to elapse between interventions or any handling
of the infant. Nothing was done before or during the interventions to induce
sleeping or crying in the infant, an inconvenient, though necessary precaution,
rarely taken in experimental work with infants. This was done, partly to avoid
differential changes in state in response to such manipulations, partly to keep
constant the amount of stimulation given each child.
The interventions were:
1. The infant was lifted from the examining table and put to the shoul-
der, head supported.
2. The infant was lifted horizontally and cradled in arms in the nursing
position.
3. Bending over the supine infant, tne assistant held the infant close,
simulating another nursing position. Care was taken not to move the
infant in any way.
4. The infant who had previously been placed in an infant seat, was
raised to the upright position.
5. The infant in the infant seat was moved horizontally, as if in a per-
ambulator.
6. The infant was talked to in a high pitched voice, simulating "mother
talk". The voice was used as a marker to avoid bias in picking a time
during which to observe alerting in the absence of any intervention.
This was decided after a preliminary study showed that the voice elic-
18
18
ited no more alerting than occurred spontaneously without any
intervention.
Table 1 shows the types of stimulation given in each intervention.
Insert Table 1 about here
Figure 1 shows in descending order the level of alertness evoked by each
intervention, both in crying and in sleeping infants.
Insert Figure 1 about here
As can be seen from Figure 1, by far the highest level of alerting was evoked
by putting the infant to the shoulder. This position evoked bright-eyed scanning
in 77.5% of the infants. The photograph in slide 3 portrays a typical reaction on
the part of a newborn on being soothed on the shoulder.
Insert photograph of infant here
Returning to Figure 1, we assessed the significance of the difference of re-
sponse to the six interventions through t tests. This was possible only for the
crying infants in as much as the minimal response of the sleeping babies made such
testing inappropriate. I shall confine myself here only to our main findings which
indicated that vestibular stimulation had a far more powerful effect in evoking
visual alertness than did contact, and that contact alone had no greater effect
than hearing a highpitched voice, which in turn, judging from preliminary work,
had no greater effect on alerting than would occur by chance. The results also
clearly showed, that even though sleeping infants responded only minimally, the only
two interventions which elicited any visual activity were interventions entailing
19
19
vestibular stimulation and the upright position.
We drew several conclusions from this study. One of these was that mothers,
in soothing their crying infants by picking them up, will inadvertently provide
them with a great many visual experiences (See also Korner and Grobstein, 1966).
Another was that if the earliest forms of learning occur mostly through visual
exploration, that vestibular stimulation which evokes a good deal of visual alert-
ness in the neonate, may be the more important form of stimulation during this stage
of development than the more publicized body contact. What is interesting is that
it is a hidden form of stimulation, easily overlooked, because it almost invariably
attends experiences involving body contact and tactile stimulation. We also con-
cluded from reading some of the embryological literature, that it makes good sense
for the vestibular system to be an excellent mediator for early stimulation. It
appears that this system is one of the earliest to develop, begins to function and
to be myelinated at 4 months gestational age and is fully mature at birth (Humphrey,
1965; Langworthy, 1933).
Returning to our topic, how does this study illustrate that state can be a medi-
ator of stimulation? Not only does the crying state evoke maternal stimulation, and
the state of visual alertness provide environmental stimulation, but depending on
the infant's state at the time of intervention, the same type of stimulation will
either have a strong or a minimal effect. This was clearly shown by the differen-
tial response of sleeping and of crying infants. The whole study suggests that in
infant research involving early stimulation, it is important to pin down not only
which types of stimulation are most effective at which stages of neurophysiological
development, but also to spell out the optimal state of the organism which will
facilitate the effectiveness of stimulation.
20
20
Summary
In this paper, the influence of several sleeping and waking
states and of hunger on the neonate's behavior are described, State
is an obstacle and confounds observations in infant research if it is
ignored. Without controlling for state, infant studies will yield
both false positive and false negative results. Differences among in-
fants will be found which are purely a function of their momentary
states rather than their individuality. Likewise, findings over ba-
bies, or ,,eneral effects of interventions will be obscured unless the
infants are observed in comparable states. If, on the other hand, one
routinely cmtrols for the infant's state, it ceases to be a source of
error in making systematic observations. Instead, state becomes a val-
uable orq,anizer of otherwise confusing data.
State as a variable in its on ri:;ht was illustrated from examples
from sleep research. Also discussed were distinctness of state, range,
predictability and alterability of states as these may affect the mo-
ther-infant relationship as well as normal and pathological development.,
The states of crying and of alert inactivity were discussed both under
the heading of state as variable and as mediator of stimulation. The
paper concludes with a discussion of the results of a study which il-
lustrates that in studies assessing the effects of early stimulation,
it is important to pin down not only which types of stimulation are
most effective at which stages of development, but also to spell out
the optimal state of the infant during which such stimulation can take
an effect.
23.
21
BIBLIOGRAPHY
Boismier,J. D., Chappell, P. F. & Meier, G. W. A behavioral inventory for
assessing states of arousal in the human newborn. Faper presented at
the Southeastern Psychological Association, Louisville, Kentucky,
April 1970.
Brazelton, T. B. Observations of the neonate. Journal of the American
Academy of Child Psychiatry, 1962, 1, 38-58.
Brazelton, T. B. From Brazelton's Newborn Scale. Personal communication,
October 19, 1970.
Brazelton, T. B., Scholl, M. L. & Robey, J. S. Visual responses in the
newborn. Pediatrics, 1966, 37, February, 284-290.
Bridger, W. H. & Birns, B. Neonates' behavioral and autonomic responses
to stress during soothing. Recent Advances in Biological Psychiatry,
1963, 5, 1-6.
Brown, J. L. States in newborn infants. herrill-Palmer Quarterly, 1964, 10,
313-327.
Bruch, H. Neuro-psychological disturbances in obesity. Psychiatric Digest,
1966, 26, 37-41.
Connolly, K. & Stratton, P. An exploration of some parameters affecting
classical conditioning in the neonate. Child Development, 1969, 40,
431-441.
Dowling, S. Three children with esophageal atresia: a longitudinal study
of the effects of hunger and oral feeding on infant development. Paper
presented at the American Psychoanalytic Association Meetings, San
Francisco, California, May 1970.
Eisenberg, R., Griffin, E. J., Coursin, D. B. & Hunter, M. A. Auditory
behavior in the human neonate: a preliminary report. Journal of Speech
and Hearing Research, 1964, 7, 245-269.
22
31.=.11117.12.......
Ellingson, R. J, The study of brain electrical activity in infants. In
L. P. Lipsitt E C, C, Spiker (Eds.), Advances in Child Development and
Behavior, Vol. 3, Now York, London: Academic Press, 1967, 93-97.
Erode, R. N, & Metcalf, D, R, An electroencephalographic study of the be-
havioral rapid eye movement states in the human newborn. Journal of
Nervous and Mental Disease, 1970, 150, 376-386.
Escalona, S. K, The study of individual diff..arences and the problem of
state. Journal of the American Academy oT Child Psychiatry, 1962, 1,
11-37.
Fisher, C., Gross. J. & Zuch, J. A cycle of penile erections synchronous
wth dreaming (REM) sleep. Archives of General Psychiatry, 1965, 12,
29-45.
Fisichelli, V. R, Haber, A., Davis, J. & Karelitz, S. Audible character-
istics of the cries of normal infants and those with Down's syndrome.
Perceptual and Motor Skills, 19E6, 23, 744-746.
Griggs, R, Co & Stunkard, A. The interpretation of gastric motility.
Sensitivity and Hat- in the perception of gastric motility. Archives
of General Psychiatry, 1964, 11, 82-89.
Hamburger, V. The concept of 'development' In biology. In D. B. Harris
(Ed.), The Concept of Develonment. Minneapolis: University of Minnesota
Press. 1947, pp. 49-58,
Humphrey, T. The embryologic differentiation, of the vestibular nuclei in
Iron correlated with functional development. International Symposiur.
on Vestibular and Oculomotor Problems, Tokyo, 1965, pp. 51-56.
Jouvet-Mounier, D. Ontogenese des Etats de Vigilance chez quelques Mammi-
feres. Lyon: J. Tixier & Fils, 1968.
Koresko, R. L,, Snyder, F. & Feinberg, I. 'Dream time' in hallucinating and
ron-hallucinating schizophrenic patients. Nature, 1963, 199, No. 4898,
23
23
1118-1119.
Korner, A. F. REM organization in neonates. Theoretical implications for
development and the biological function of REM. Archives of General
Psychiatry, 1968, 19, 330-340.
Korner, A. F, Neonatal startles, smiles, erections, and reflex sucks as
related to state, sex, and individuality. Child Development, 1969, 40,
No. 4, 1039-1053,
Korner, A. F. Visual alertness in neonates: Individual differences and their
correlates. Perceptual and Motor Skills, 1970, 31, 499-509.
Korner, A. F., Chuck, B. & Dontchos, S. Organismic determinants of spon-
taneous oral behavior in neonates. Child Development, 1968, 12, No. 4,
1145-1157.
Korner, A. F. & Grobstein, R. Visual alertness as related to soothing in
neonates: implications for maternal stimulation and early deprivation.
Child Development, 1966, 37, 867-876.
Korner, A. F. & Grobstein, R. Individual differences at birth: implications
for mother-infant relationship and later development. Journal of the
American Academy of Child Psychiatry, 1967, 6, No. 4, 676-690.
Korner, A. F. & Kraemer, N. Co Individual differences in spontaneous oral
behavior. Third Symposium on Oral Sensation and Perception: The Mouth
of the Infant, J. F. Bosma (Ed.), Springfield, Charles C. Thomas,
1971, in press.
Korner, A. F. & Thoman, E. B. Visual alertness in neonates as evoked by
maternal care. Journal of Experimental Child Psychology, 1970, 10,
67-78.
Langworthy, O. R. Development of behavior patterns and myelinization of
the nervous system in the human fetus and infant. Carnegie Institution
of Washington Contributions to Embryology, 1933, 24, 1-57.
24
24
Lenard, H. G., von Bernuth, H. & Prechtl, H. F. R. Reflexes and their
relationship to behavioural state in the newborn. Acta Paediatrica
Scandinavica, 1968, 52, 177-185.
Monod, N., Eliet-Flescher, J. & Dreyfus-Bris,ae, C. The sleep of the
full-term newborn and premature infant. Biologic Neonatorium, 1967,
11, 216-247.
Parmelee, A. H., Wenner, W. H., Akiyama, Y., Schultz, M. & Stern, E. Sleep
states in premature infants. Developmental Medicine and Child Neurology,
1967, 9, 70-77.
Prechtl, H. F. R., Akiyama, Y., Zinkin, P. & Grant, D. K. Polygraphic studies
of the full-term newborn: I. Technical aspects and qualitative analysis.
In: Studies in Infancy. Clinics in Develouental Medicine, 1968, 22, 1-21.
Prechtl, H. F. R., & Beintema, D. The neurological examination of the full-
term newborn infant. Little Club Clinics in Developmental Medicine, 1964,
12, 1-74.
Prechtl, H. F, R., Theorell, K., Gramsbergen, A. & Lind, J. A statistical
analysis of cry patterns in normal and abnormal newborn infants. Develop-
mental Medicine and Child Neurology, 1969, 11, 142-152.
Roffwarg, H. P., Muzio, J. N. & Dement, W. C. Ontogenetic development of the
human sleep-dream cycle. Science, 1966, 152, 604-619.
Vincent, J. D., Favarel-Garrigues, B., Bourgeois, M. & Dufet, B. Sommeil de
nuit du schizophrene en debut devolution. Etude polygraphique par teleme-
sure. Annales Medico-Psycholoqiques, 1968, 126, 227-236.
Weller, G. M. & Bell, R. Q. Basal skin conductance and neonatal state. Child
Development, 1965, 16, 647-657.
Werner, H. The concept of development from a comparative and organismic point
of view. In D. B. Harris (Ed.), The Concept of Development. Minneapolis:
University of Minnesota Press, 1947, pp. 125-148.
25
25
Williams, H. L., Hammack, J. T., Daly, R. L., Dement, W. C. & Lubin, A.
Responses to auditory stimulation, sleep loss and the EEG stages of
sleep. Electroencephalography and Clinical Neurophysiology, 1964,
16, 269-279.
Wolff, P. H. Observations on newborn infants. Psychosomatic Medicine, 1959,
21,, 110-118.
Wolff, P. H. The causes, controls, and organization of behavior in the
neonate. Psychological Issues, 1966, 5, No. 1, Monograph 17.
26
26
* Department of Psychiatry, Stanford University School of Medicine,
Stanford, California 94305
1
Preparation of this manuscript was facilitated by the support of
USPHS Grant HD- 3591 from the National Institute of Child Health and
Human Development. This paper was presented as part of a Symposium on
"The Concept of State: Its Relevance to infant Behavior" on February 12,
1971, held at the Merrill-Palmer Institute Conference on Research and
Teaching of Infant Development.
27
27
TABLE 1
TYPES OF STIMULATION PROVIDED
BY EACH INTERVENTION
Interventions Stimulation
I. To shoulder C V U
2. To breast C V -
3. Held close C - -
4. Infant seat up - V U
5. Infant seat to side - V -
6. Female voice
Stimlation: C = Contact
V = Vestibular
U = Upright
28
28
VISUAL
ALERTNESS
SCORES
Effects of Stim
uIolion01.1\11.stao,1 A
lerting
Shoulder sidle -side,S
it - up To breosi E
mbrace
Voice
c\iV
VU
CV
MEAN VISUAL ALERTNESS SCORES FOR
64 TWO DAY OLD NEONATES IN RESPONSE
TO 6 STIMULUS CONFIGURATIONS